mailing list archives
Re: Nmap script ideas wiki
From: David Fifield <david () bamsoftware com>
Date: Fri, 11 Mar 2011 09:48:57 -0800
On Fri, Mar 11, 2011 at 07:13:14PM +0200, Toni Ruottu wrote:
What kind of suggestions are welcome? I wrote a few, but I am not sure
if I should spam the wiki with all the scripts I've been thinking
about. On the other hand some of the topics may be useful. Even if
they just provide new comers a list of things they can choose from.
There are some info script categories which may easily get long, and
may not need exact descriptions, as the purpose of an info script is
always just to extract some information semitrivially available
through chatting the protocol. Some types I have been looking at
game servers (quake3-info, wesnoth-info)
These reveal information about the game world, but also technical
information about the server configuration. They are often simple to
write, but may require parsing all kinds of formats that may or may
not be trivial. quake3 style servers require talking a binary
protocol, while wesnoth uses gzip and xml.
network diagnostic services (teredo-info, stun-info)
These may reveal lots of information about the targets, but also some
information about the network environment between the scanner and the
target. These scripts have lots of potential, but may be hard to write
as there is a lot one could do. Also writing these requires lots of
rfc reading, as the specifications may be long (teredo) or scattered
in multiple rfcs (stun, turn, ice, ...)
system monitoring services (gkrellm-info, mbmon-info)
These are really good targets for script writing, as the services are
designed to reveal lots of interesting information about the system.
The produced scripts are also really useful for administrators, as
they can then use nmap for gathering statistics of multiple machines
with nmap scans. The problem with these is that available information
may be overwhelming. For example gkrellm reports the cpu load with an
interval of a few seconds. What should the script show to the user? A
graph? Average value? First value? Min and max values? Ofcourse there
is lots of simple information available as well, but deciding what to
show and how may be hard.
remote administration tools (backorifice-info, subseven-info,
These are important because insecure remote administration tools may
reveal lots of information about the system. It is critical to
acknowledge any such services as soon as possible. Most of the ones I
listed above are used by malicious users to gain access of
unsuspecting victims, so highlighting these systems to the admin is
really useful for improving security. Most of these are old, but some
of them still work with up to date systems. The problem with these is
that the protocols may not be clearly documented, so one needs to do
research with wireshark, and google to find out how they work.
Grepping open source reimplementations is also useful.
peer-to-peer nodes (gnutella-info, tor-info, freenet-info)
Peer-to-peer nodes often publish technical information to co-operate
with other nodes. Having convenient access to this information is
useful for researching the system, but also to give users some idea
what kind of data they are giving out to the world. The problems
involved with these are that there may be lots of information
available, so one needs to decide what to show to the user. Some
information may also be relative to your position in the network. Some
of these services reveal a connection table, which makes it possible
to draw graphs about the systems, or crawl the network to scan other
nodes involved in the protocol.
discovery services (udp-bittorrenttracker-info,
These scripts are useful for getting some nodes to scan while
exploring a peer-to-peer system. They can provide a starting point for
crawling the network. The discovery services may also provide other
interesting information. Also, getting a list of IP addresses when
ever the scan hits a discovery server makes it clear to the user what
the services is used for. There are two types of discovery scripts.
Some have a pre rule and are mainly used to choose scan targets for a
scan, but some other are used by scanning the discovery service to
extract information out of it. Also the latter ones may be used to get
scanning targets, but this typically leads to scanning the discovered
servers for discovery services, which is a bit odd.
These examples are from the top of my head. I just thought I'd post
them here rather than spam the wiki directly. We can always move some
of these to the wiki, if that is useful.
I think these are fine examples of what we want on the page. I would
copy all of them there. One of the motivations behind putting ideas on
the wiki is make discussion more permanent, and to make it easier for
other people to contribute (who may see joining a mailing list as too
high a barrier). We can also put notes next to entries to show which
ones are being worked on.
In general, I'd say use the Wikipedia philosphy and Be Bold in editing.
Sent through the nmap-dev mailing list
Archived at http://seclists.org/nmap-dev/
Re: Nmap script ideas wiki Ron (Mar 14)